CRO, short for conversion rate optimization, is the work of figuring out why visitors who came to your store did not buy, and what to change so more of them do. Instead of paying for more traffic, CRO tries to get more out of the traffic that already arrives, so every improvement raises the return of every marketing channel at once. In practice it is a systematic process, not a list of tips: measure the funnel, find the step where the most visitors fall away, form a hypothesis, and test it against the current version. Below: what CRO is, how the process works, a worked example, and the mistakes that turn tests into noise.
What is CRO
CRO (conversion rate optimization) is an ongoing process of measuring and testing, aimed at getting more visitors to buy from the same traffic.
- Full name: conversion rate optimization.
- The metric it tries to move: conversion rate.
- What it is not: just a redesign. A simpler checkout, visible shipping cost and faster page loads are part of it too.
How CRO works
- Measure the funnel. Product view, add to cart, checkout start and purchase. At each step, record how many visitors moved on.
- Diagnose. Look at analytics, session recordings, short surveys, and the questions customers ask before buying.
- Form a hypothesis. For example: "visitors leave the product page because shipping cost is unclear".
- Test. An A/B test against the current version, with sample size and duration fixed in advance.
- Ship and document. Even a losing test teaches something, provided you wrote it down.
Example: shipping cost on the product page
- A store gets 10,000 sessions a month: 4,000 view a product, 800 add to cart, 400 start checkout and 200 buy. Conversion rate: 2%.
- The largest relative drop is between viewing and adding to cart: only 20% move on.
- The store starts showing shipping cost on the product page.
- Add-to-cart rises to 22%, and the later steps hold their ratios.
- The result: 880 adds, 440 checkouts and 220 purchases, ten percent more orders from the same traffic.
The figures in this article's examples are illustrative only and do not reflect actual customer data.
CRO for an online store
- Cheaper customers: the same ad spend brings more buyers.
- Compounding gains: more traffic or a higher order value is worth more when the store converts better.
- Where to start: practical ideas are collected in the guide to increasing ecommerce conversion rate.
Common mistakes
- Copying "best practices" without data. What worked for another store may not fit your audience.
- Stopping a test the moment it looks good. Repeated peeking and early stopping produce wins that later vanish.
- Testing tiny changes on little traffic. A small store is better off testing bold changes that can show an effect.
- Improving one step at the business's expense. A discount popup can lift conversion and train customers to wait for a discount.
- Forgetting mobile. Most tests are designed on a big screen while the traffic arrives on phones.
Related metrics
- Conversion rate: the metric CRO tries to move.
- AOV: should hold steady, so the gain does not come out of order size.
- Cart abandonment and page exit rate: show where to look.
CRO in StoreChart
- Not included: StoreChart does not run A/B tests and does not measure site sessions or funnel steps. Those come from your analytics tool.
- The AI chatbot: answers visitors from a knowledge base built from the store's pages, products and FAQs, and can recommend a product.
- Saved conversations: reading them shows which questions visitors ask before they buy, good raw material for CRO hypotheses.
Frequently asked questions
Where should a store start with CRO?
By finding the step in the funnel with the sharpest drop-off, analytics showing where most visitors abandon (product page, cart, checkout) points directly to which fix will have the biggest impact, rather than guessing at page copy first.
How long should an A/B test run?
Long enough to reach statistical significance, which depends on traffic volume and the size of the effect being tested, ending a test after just a few days or a small sample size risks drawing conclusions from random noise rather than a real difference. Running for at least one full week is also common, to cover day-of-week differences.
Is CRO only about redesigning pages?
No, CRO also includes things like simplifying checkout steps, clarifying shipping costs earlier, adding trust signals (reviews, guarantees), and improving page load speed, all of which can move conversion rate without a visual redesign.
